Adversarial attacks constitute a notable threat to machine learning systems, given their potential to induce erroneous predictions and classifications. However, within real-world contexts, the essential specifics of the deployed model are frequently treated as a black box, consequently mitigating the vulnerability to such attacks. Thus, enhancing the transferability of the adversarial samples has become a crucial area of research, which heavily relies on selecting appropriate surrogate models. To address this challenge, we propose a novel approach that generates adversarial attacks in a mutual-modality optimization scheme. Our approach is accomplished by leveraging the pre-trained CLIP model. Firstly, we conduct a visual attack on the clean image that causes semantic perturbations on the aligned embedding space with the other textual modality. Then, we apply the corresponding defense on the textual modality by updating the prompts, which forces the re-matching on the perturbed embedding space. Finally, to enhance the attack transferability, we utilize the iterative training strategy on the visual attack and the textual defense, where the two processes optimize from each other. We evaluate our approach on several benchmark datasets and demonstrate that our mutual-modal attack strategy can effectively produce high-transferable attacks, which are stable regardless of the target networks. Our approach outperforms state-of-the-art attack methods and can be readily deployed as a plug-and-play solution.
论文检索
输入标题、作者或关键词，从 7,876 篇学术成果中精准定位
Supervised Contrastive Loss (SCL) is popular in visual representation learning. Given an anchor image, SCL pulls two types of positive samples, i.e., its augmentation and other images from the same class together, while pushes negative images apart to optimize the learned embedding. In the scenario of long-tailed recognition, where the number of samples in each class is imbalanced, treating two types of positive samples equally leads to the biased optimization for intra-category distance. In addition, similarity relationship among negative samples, that are ignored by SCL, also presents meaningful semantic cues. To improve the performance on long-tailed recognition, this paper addresses those two issues of SCL by decoupling the training objective. Specifically, it decouples two types of positives in SCL and optimizes their relations toward different objectives to alleviate the influence of the imbalanced dataset. We further propose a patch-based self distillation to transfer knowledge from head to tail classes to relieve the under-representation of tail classes. It uses patch-based features to mine shared visual patterns among different instances and leverages a self distillation procedure to transfer such knowledge. Experiments on different long-tailed classification benchmarks demonstrate the superiority of our method. For instance, it achieves the 57.7% top-1 accuracy on the ImageNet-LT dataset. Combined with the ensemble-based method, the performance can be further boosted to 59.7%, which substantially outperforms many recent works. Our code will be released.
Recently, CLIP has found practical utility in the domain of pixel-level zero-shot segmentation tasks. The present landscape features two-stage methodologies beset by issues such as intricate pipelines and elevated computational costs. While current one-stage approaches alleviate these concerns and incorporate Visual Prompt Training (VPT) to uphold CLIP's generalization capacity, they still fall short in fully harnessing CLIP's potential for pixel-level unseen class demarcation and precise pixel predictions. To further stimulate CLIP's zero-shot dense prediction capability, we propose SPT-SEG, a one-stage approach that improves CLIP's adaptability from image to pixel. Specifically, we initially introduce Spectral Prompt Tuning (SPT), incorporating spectral prompts into the CLIP visual encoder's shallow layers to capture structural intricacies of images, thereby enhancing comprehension of unseen classes. Subsequently, we introduce the Spectral Guided Decoder (SGD), utilizing both high and low-frequency information to steer the network's spatial focus towards more prominent classification features, enabling precise pixel-level prediction outcomes. Through extensive experiments on two public datasets, we demonstrate the superiority of our method over state-of-the-art approaches, performing well across all classes and particularly excelling in handling unseen classes.
MuLTI: Efficient Video-and-Language Understanding with Text-Guided MultiWay-Sampler and Multiple Choice Modeling
PDF ↗Video-and-language understanding has a variety of applications in the industry, such as video question answering, text-video retrieval, and multi-label classification. Existing video-and-language understanding methods generally adopt heavy multi-modal encoders and feature fusion modules, which consume high computational costs. Specially, they have difficulty dealing with dense video frames or long text prevalent in industrial applications. This paper proposes MuLTI, a highly accurate and efficient video-and-language understanding model that achieves efficient and effective feature fusion and rapid adaptation to downstream tasks. Specifically, we design a Text-Guided MultiWay-Sampler based on adapt-pooling residual mapping and self-attention modules to sample long sequences and fuse multi-modal features, which reduces the computational costs and addresses performance degradation caused by previous samplers. Therefore, MuLTI can handle longer sequences with limited computational costs. Then, to further enhance the model's performance and fill in the lack of pretraining tasks in the video question answering, we propose a new pretraining task named Multiple Choice Modeling. This task bridges the gap between pretraining and downstream tasks and improves the model's ability to align video and text features. Benefiting from the efficient feature fusion module and the new pretraining task, MuLTI achieves state-of-the-art performance on multiple datasets. Implementation and pretrained models will be released.
The capacity to generalize to future unseen data stands as one of the utmost crucial attributes of deep neural networks. Sharpness-Aware Minimization (SAM) aims to enhance the generalizability by minimizing worst-case loss using one-step gradient ascent as an approximation. However, as training progresses, the non-linearity of the loss landscape increases, rendering one-step gradient ascent less effective. On the other hand, multi-step gradient ascent will incur higher training cost. In this paper, we introduce a normalized Hessian trace to accurately measure the curvature of loss landscape on both training and test sets. In particular, to counter excessive non-linearity of loss landscape, we propose Curvature Regularized SAM (CR-SAM), integrating the normalized Hessian trace as a SAM regularizer. Additionally, we present an efficient way to compute the trace via finite differences with parallelism. Our theoretical analysis based on PAC-Bayes bounds establishes the regularizer's efficacy in reducing generalization error. Empirical evaluation on CIFAR and ImageNet datasets shows that CR-SAM consistently enhances classification performance for ResNet and Vision Transformer (ViT) models across various datasets. Our code is available at https://github.com/TrustAIoT/CR-SAM.
Spiking neural networks (SNNs) have revolutionized neural learning and are making remarkable strides in image analysis and robot control tasks with ultra-low power consumption advantages. Inspired by this success, we investigate the application of spiking neural networks to 3D point cloud processing. We present a point-to-spike residual learning network for point cloud classification, which operates on points with binary spikes rather than floating-point numbers. Specifically, we first design a spatial-aware kernel point spiking neuron to relate spiking generation to point position in 3D space. On this basis, we then design a 3D spiking residual block for effective feature learning based on spike sequences. By stacking the 3D spiking residual blocks, we build the point-to-spike residual classification network, which achieves low computation cost and low accuracy loss on two benchmark datasets, ModelNet40 and ScanObjectNN. Moreover, the classifier strikes a good balance between classification accuracy and biological characteristics, allowing us to explore the deployment of 3D processing to neuromorphic chips for developing energy-efficient 3D robotic perception systems.
Video Person Re-Identification (Re-ID) is a task of retrieving persons from multi-camera surveillance systems. Despite the progress made in leveraging spatio-temporal information in videos, occlusion in dense crowds still hinders further progress. To address this issue, we propose a Temporal Correlation Vision Transformer (TCViT) for video person Re-ID. TCViT consists of a Temporal Correlation Attention (TCA) module and a Learnable Temporal Aggregation (LTA) module. The TCA module is designed to reduce the impact of non-target persons by relative state, while the LTA module is used to aggregate frame-level features based on their completeness. Specifically, TCA is a parameter-free module that first aligns frame-level features to restore semantic coherence in videos and then enhances the features of the target person according to temporal correlation. Additionally, unlike previous methods that treat each frame equally with a pooling layer, LTA introduces a lightweight learnable module to weigh and aggregate frame-level features under the guidance of a classification score. Extensive experiments on four prevalent benchmarks demonstrate that our method achieves state-of-the-art performance in video Re-ID.
The recent contrastive language-image pre-training (CLIP) model has shown great success in a wide range of image-level tasks, revealing remarkable ability for learning powerful visual representations with rich semantics. An open and worthwhile problem is efficiently adapting such a strong model to the video domain and designing a robust video anomaly detector. In this work, we propose VadCLIP, a new paradigm for weakly supervised video anomaly detection (WSVAD) by leveraging the frozen CLIP model directly without any pre-training and fine-tuning process. Unlike current works that directly feed extracted features into the weakly supervised classifier for frame-level binary classification, VadCLIP makes full use of fine-grained associations between vision and language on the strength of CLIP and involves dual branch. One branch simply utilizes visual features for coarse-grained binary classification, while the other fully leverages the fine-grained language-image alignment. With the benefit of dual branch, VadCLIP achieves both coarse-grained and fine-grained video anomaly detection by transferring pre-trained knowledge from CLIP to WSVAD task. We conduct extensive experiments on two commonly-used benchmarks, demonstrating that VadCLIP achieves the best performance on both coarse-grained and fine-grained WSVAD, surpassing the state-of-the-art methods by a large margin. Specifically, VadCLIP achieves 84.51% AP and 88.02% AUC on XD-Violence and UCF-Crime, respectively. Code and features are released at https://github.com/nwpu-zxr/VadCLIP.
Multiscale Low-Frequency Memory Network for Improved Feature Extraction in Convolutional Neural Networks
PDF ↗Deep learning and Convolutional Neural Networks (CNNs) have driven major transformations in diverse research areas. However, their limitations in handling low-frequency in-formation present obstacles in certain tasks like interpreting global structures or managing smooth transition images. Despite the promising performance of transformer struc-tures in numerous tasks, their intricate optimization com-plexities highlight the persistent need for refined CNN en-hancements using limited resources. Responding to these complexities, we introduce a novel framework, the Mul-tiscale Low-Frequency Memory (MLFM) Network, with the goal to harness the full potential of CNNs while keep-ing their complexity unchanged. The MLFM efficiently preserves low-frequency information, enhancing perfor-mance in targeted computer vision tasks. Central to our MLFM is the Low-Frequency Memory Unit (LFMU), which stores various low-frequency data and forms a parallel channel to the core network. A key advantage of MLFM is its seamless compatibility with various prevalent networks, requiring no alterations to their original core structure. Testing on ImageNet demonstrated substantial accuracy improvements in multiple 2D CNNs, including ResNet, MobileNet, EfficientNet, and ConvNeXt. Furthermore, we showcase MLFM's versatility beyond traditional image classification by successfully integrating it into image-to-image translation tasks, specifically in semantic segmenta-tion networks like FCN and U-Net. In conclusion, our work signifies a pivotal stride in the journey of optimizing the ef-ficacy and efficiency of CNNs with limited resources. This research builds upon the existing CNN foundations and paves the way for future advancements in computer vision. Our codes are available at https://github.com/AlphaWuSeu/MLFM.
Remote sensing imagery, despite its broad applications in helping achieve Sustainable Development Goals and tackle climate change, has not yet benefited from the recent advancements of versatile, task-agnostic vision language models (VLMs). A key reason is that the large-scale, semantically diverse image-text dataset required for developing VLMs is still absent for remote sensing images. Unlike natural images, remote sensing images and their associated text descriptions cannot be efficiently collected from the public Internet at scale. In this work, we bridge this gap by using geo-coordinates to automatically connect open, unlabeled remote sensing images with rich semantics covered in OpenStreetMap, and thus construct SkyScript, a comprehensive vision-language dataset for remote sensing images, comprising 2.6 million image-text pairs covering 29K distinct semantic tags. With continual pre-training on this dataset, we obtain a VLM that surpasses baseline models with a 6.2% average accuracy gain in zero-shot scene classification across seven benchmark datasets. It also demonstrates the ability of zero-shot transfer for fine-grained object attribute classification and cross-modal retrieval. We hope this dataset can support the advancement of VLMs for various multi-modal tasks in remote sensing, such as open-vocabulary classification, retrieval, captioning, and text-to-image synthesis.
Open set recognition (OSR) requires the model to classify samples that belong to closed sets while rejecting unknown samples during test. Currently, generative models often perform better than discriminative models in OSR, but recent studies show that generative models may be computationally infeasible or unstable on complex tasks. In this paper, we provide insights into OSR and find that learning supplementary representations can theoretically reduce the open space risk. Based on the analysis, we propose a new model, namely Multi-Expert Diverse Attention Fusion (MEDAF), that learns diverse representations in a discriminative way. MEDAF consists of multiple experts that are learned with an attention diversity regularization term to ensure the attention maps are mutually different. The logits learned by each expert are adaptively fused and used to identify the unknowns through the score function. We show that the differences in attention maps can lead to diverse representations so that the fused representations can well handle the open space. Extensive experiments are conducted on standard and OSR large-scale benchmarks. Results show that the proposed discriminative method can outperform existing generative models by up to 9.5% on AUROC and achieve new state-of-the-art performance with little computational cost. Our method can also seamlessly integrate existing classification models. Code is available at https://github.com/Vanixxz/MEDAF.
Data augmentation is an effective regularization strategy for mitigating overfitting in deep neural networks, and it plays a crucial role in 3D vision tasks, where the point cloud data is relatively limited. While mixing-based augmentation has shown promise for point clouds, previous methods mix point clouds either on block level or point level, which has constrained their ability to strike a balance between generating diverse training samples and preserving the local characteristics of point clouds. The significance of each part component of the point clouds has not been fully considered, as not all parts contribute equally to the classification task, and some parts may contain unimportant or redundant information. To overcome these challenges, we propose PointPatchMix, a novel approach that mixes point clouds at the patch level and integrates a patch scoring module to generate content-based targets for mixed point clouds. Our approach preserves local features at the patch level, while the patch scoring module assigns targets based on the content-based significance score from a pre-trained teacher model. We evaluate PointPatchMix on two benchmark datasets including ModelNet40 and ScanObjectNN, and demonstrate significant improvements over various baselines in both synthetic and real-world datasets, as well as few-shot settings. With Point-MAE as our baseline, our model surpasses previous methods by a significant margin. Furthermore, our approach shows strong generalization across various point cloud methods and enhances the robustness of the baseline model. Code is available at https://jiazewang.com/projects/pointpatchmix.html.
Understanding vehicles in images is important for various applications such as intelligent transportation and self-driving system. Existing vehicle-centric works typically pre-train models on large-scale classification datasets and then fine-tune them for specific downstream tasks. However, they neglect the specific characteristics of vehicle perception in different tasks and might thus lead to sub-optimal performance. To address this issue, we propose a novel vehicle-centric pre-training framework called VehicleMAE, which incorporates the structural information including the spatial structure from vehicle profile information and the semantic structure from informative high-level natural language descriptions for effective masked vehicle appearance reconstruction. To be specific, we explicitly extract the sketch lines of vehicles as a form of the spatial structure to guide vehicle reconstruction. The more comprehensive knowledge distilled from the CLIP big model based on the similarity between the paired/unpaired vehicle image-text sample is further taken into consideration to help achieve a better understanding of vehicles. A large-scale dataset is built to pre-train our model, termed Autobot1M, which contains about 1M vehicle images and 12693 text information. Extensive experiments on four vehicle-based downstream tasks fully validated the effectiveness of our VehicleMAE. The source code and pre-trained models will be released at https://github.com/Event-AHU/VehicleMAE.
The main streams of human activity recognition (HAR) algorithms are developed based on RGB cameras which usually suffer from illumination, fast motion, privacy preservation, and large energy consumption. Meanwhile, the biologically inspired event cameras attracted great interest due to their unique features, such as high dynamic range, dense temporal but sparse spatial resolution, low latency, low power, etc. As it is a newly arising sensor, even there is no realistic large-scale dataset for HAR. Considering its great practical value, in this paper, we propose a large-scale benchmark dataset to bridge this gap, termed HARDVS, which contains 300 categories and more than 100K event sequences. We evaluate and report the performance of multiple popular HAR algorithms, which provide extensive baselines for future works to compare. More importantly, we propose a novel spatial-temporal feature learning and fusion framework, termed ESTF, for event stream based human activity recognition. It first projects the event streams into spatial and temporal embeddings using StemNet, then, encodes and fuses the dual-view representations using Transformer networks. Finally, the dual features are concatenated and fed into a classification head for activity prediction. Extensive experiments on multiple datasets fully validated the effectiveness of our model. Both the dataset and source code will be released at https://github.com/Event-AHU/HARDVS.
The Novel Category Discovery problem aims to cluster an unlabeled set with the help of a labeled set consisting of disjoint but related classes. However, existing models treat class names as discrete one-hot labels and ignore the semantic understanding of these classes. In this paper, we propose a new setting named Semantic-guided Novel Category Discovery (SNCD), which requires the model to not only cluster the unlabeled images but also semantically recognize these images based on a set of their class names. The first challenge we confront pertains to effectively leveraging the class names of unlabeled images, given the inherent gap between the visual and linguistic domains. To address this issue, we incorporate a semantic-aware recognition mechanism. This is achieved by constructing dynamic class-wise visual prototypes as well as a semantic similarity matrix that enables the projection of visual features into the semantic space. The second challenge originates from the granularity disparity between the classification and clustering tasks. To deal with this, we develop a semantic-aware clustering process to facilitate the exchange of knowledge between the two tasks. Through extensive experiments, we demonstrate the mutual benefits of the recognition and clustering tasks, which can be jointly optimized. Experimental results on multiple datasets confirm the effectiveness of our proposed method. Our code is available at https://github.com/wang-weishuai/Semantic-guided-NCD.
Detecting in open-world scenarios poses a formidable challenge for models intended for real-world deployment. The advanced closed set object detectors achieve impressive performance under the closed set setting, but often produce overconfident misprediction on unknown objects due to the lack of supervision. In this paper, we propose a novel Evidential Object Detector (EOD) to formulate the Open Set Object Detection (OSOD) problem from the perspective of Evidential Deep Learning (EDL) theory, which quantifies classification uncertainty by placing the Dirichlet Prior over the categorical distribution parameters. The task-specific customized evidential framework, equipped with meticulously designed model architecture and loss function, effectively bridges the gap between EDL theory and detection tasks. Moreover, we utilize contrastive learning as an implicit means of evidential regularization and to encourage the class separation in the latent space. Alongside, we innovatively model the background uncertainty to further improve the unknown discovery ability. Extensive experiments on benchmark datasets demonstrate the outperformance of the proposed method over existing ones.
In practical black-box attack scenarios, most of the existing transfer-based attacks employ pretrained models (e.g. ResNet50) as the substitute models. Unfortunately, these substitute models are not always appropriate for transfer-based attacks. Firstly, these models are usually trained on a largescale annotated dataset, which is extremely expensive and time-consuming to construct. Secondly, the primary goal of these models is to perform a specific task, such as image classification, which is not developed for adversarial attacks. To tackle the above issues, i.e., high cost and over-fitting on taskspecific models, we propose an Affordable and Generalizable Substitute (AGS) training framework tailored for transferbased adversarial attack. Specifically, we train the substitute model from scratch by our proposed adversary-centric constrastive learning. This proposed learning mechanism introduces another sample with slight adversarial perturbations as an additional positive view of the input image, and then encourages the adversarial view and two benign views to interact comprehensively with each other. To further boost the generalizability of the substitute model, we propose adversarial invariant learning to maintain the representations of the adversarial example invariants under augmentations with various strengths. Our AGS model can be trained solely with unlabeled and out-of domain data and avoid overfitting to any task-specific models, because of its inherently self-supervised nature. Extensive experiments demonstrate that our AGS achieves comparable or superior performance compared to substitute models pretrained on the complete ImageNet training set, when executing attacks across a diverse range of target models, including ViTs, robustly trained models, object detection and segmentation models. Our source codes are available at https://github.com/lwmming/AGS.
Recently, the rise of large-scale vision-language pretrained models like CLIP, coupled with the technology of Parameter-Efficient FineTuning (PEFT), has captured substantial attraction in video action recognition. Nevertheless, prevailing approaches tend to prioritize strong supervised performance at the expense of compromising the models' generalization capabilities during transfer. In this paper, we introduce a novel Multimodal, Multi-task CLIP adapting framework named M2-CLIP to address these challenges, preserving both high supervised performance and robust transferability. Firstly, to enhance the individual modality architectures, we introduce multimodal adapters to both the visual and text branches. Specifically, we design a novel visual TED-Adapter, that performs global Temporal Enhancement and local temporal Difference modeling to improve the temporal representation capabilities of the visual encoder. Moreover, we adopt text encoder adapters to strengthen the learning of semantic label information. Secondly, we design a multi-task decoder with a rich set of supervisory signals, including the original contrastive learning head, a cross-modal classification head, a cross-modal masked language modeling head, and a visual classification head. This multi-task decoder adeptly satisfies the need for strong supervised performance within a multimodal framework. Experimental results validate the efficacy of our approach, demonstrating exceptional performance in supervised learning while maintaining strong generalization in zero-shot scenarios.
EarthVQA: Towards Queryable Earth via Relational Reasoning-Based Remote Sensing Visual Question Answering
PDF ↗Earth vision research typically focuses on extracting geospatial object locations and categories but neglects the exploration of relations between objects and comprehensive reasoning. Based on city planning needs, we develop a multi-modal multi-task VQA dataset (EarthVQA) to advance relational reasoning-based judging, counting, and comprehensive analysis. The EarthVQA dataset contains 6000 images, corresponding semantic masks, and 208,593 QA pairs with urban and rural governance requirements embedded. As objects are the basis for complex relational reasoning, we propose a Semantic OBject Awareness framework (SOBA) to advance VQA in an object-centric way. To preserve refined spatial locations and semantics, SOBA leverages a segmentation network for object semantics generation. The object-guided attention aggregates object interior features via pseudo masks, and bidirectional cross-attention further models object external relations hierarchically. To optimize object counting, we propose a numerical difference loss that dynamically adds difference penalties, unifying the classification and regression tasks. Experimental results show that SOBA outperforms both advanced general and remote sensing methods. We believe this dataset and framework provide a strong benchmark for Earth vision's complex analysis. The project page is at https://Junjue-Wang.github.io/homepage/EarthVQA.
The success of data mixing augmentations in image classification tasks has been well-received. However, these techniques cannot be readily applied to object detection due to challenges such as spatial misalignment, foreground/background distinction, and plurality of instances. To tackle these issues, we first introduce a novel conceptual framework called Supervision Interpolation (SI), which offers a fresh perspective on interpolation-based augmentations by relaxing and generalizing Mixup. Based on SI, we propose LossMix, a simple yet versatile and effective regularization that enhances the performance and robustness of object detectors and more. Our key insight is that we can effectively regularize the training on mixed data by interpolating their loss errors instead of ground truth labels. Empirical results on the PASCAL VOC and MS COCO datasets demonstrate that LossMix can consistently outperform state-of-the-art methods widely adopted for detection. Furthermore, by jointly leveraging LossMix with unsupervised domain adaptation, we successfully improve existing approaches and set a new state of the art for cross-domain object detection.